This study will examine the performance of maximum likelihood methods of phylogeny reconstruction under a number of conditions. The approach requires simulating a large number of DNA sequence data sets, finding a maximum likelihood solution to each data set, and replicating this result over a number of data sets to in order to make statistical inferences on the methodology's performance. Our initial simulations will examine only four taxon trees; however, we are examining data that is simulated under a factorial design that results in approximately 5,600,000 data sets in the initial phases of this experiment. In the future, we intend to apply these approaches to data sets that consist of more than four taxa.